We have read the report by Sassone et al in a recent issue of the Journal. In a retrospective, single-center study, they observed that patients with left bundle branch block (LBBB) and QRS duration (QRSd) ≥178 ms who were treated by cardiac resynchronization therapy (CRT) had worse 6-month echocardiographic response and greater likelihood of adverse clinical events during a mean follow-up of 32 months compared with the rest of the LBBB population. This observation was interpreted that there is an upper limit of QRSd above which CRT benefit starts to decrease. Although the study can be viewed as a new paradigm in CRT, we would like to challenge authors’ conclusions.

The findings of the aforementioned observational study cannot be interpreted in terms of clinical benefit from CRT because a control group of conventionally treated patients is missing. Simple comparison of clinical event rates in patient subgroups by QRSd discloses no information about clinical effectiveness of CRT per se. On one side, there is enough evidence that the risk of adverse outcome in general population of patients with chronic heart failure correlates positively with QRSd. Thus, one could easily imagine that negative prognostic impact of very wide QRS may overwhelm the clinical effect of CRT. However, disregarding the cohort of CRT patients at the lower end of QRSd spectrum, none of large CRT studies demonstrated the impact of baseline QRSd on clinical outcome. Such findings indirectly suggest that intrinsic CRT effects should be greater in patients with wider QRS to compensate for their inherent risk. To further support their conclusions, the authors argue that their observation concerning very wide QRS is in line with the result of a recent meta-analysis of randomized CRT trials. We believe that such comparison is again not fully appropriate because of principally different study designs (observational vs randomized) mentioned previously. In addition, their interpretation of meta-analysis data as “declining efficacy from (QRSd) values close to 175-180 ms onward” derived from the graph of hazard ratios for the effects of CRT versus control against QRSd is apparently incorrect. On the contrary, the graph shows a progressive increase in the benefit from CRT for all-cause mortality as QRSd increased and at most plateauing in the benefit from 170 ms onward for combined end point of death or heart failure hospitalization.

Regarding 6-month echocardiographic response in patients with LBBB with QRSd ≥178 ms, which was rather low (∼23%) when defined as a ≥15% reduction of the left ventricular (LV) end-systolic volume, we may provide considerably different look at LV reverse remodeling in this subgroup of patients. We have reanalyzed data from our own cohort of CRT patients that was already investigated and published earlier in part. We were able to collect 266 patients with LBBB by Strauss’ criteria, of whom 59 patients had baseline QRSd ≥178 ms. Patients from this “high risk” subgroup had significantly higher 12-month CRT response rate (defined as a ≥10% reduction of the LV end-systolic diameter) of 73% versus 56% (p = 0.03), which translated to nonsignificantly different heart failure (10% vs 12%, p = 0.40) and cardiac mortality (15% vs 15%, p = 0.96) during the mean 3.3-year follow-up. Although such discrepancy may be partly explained by differences in the timing of echocardiographic examination and in criteria of LV reverse remodeling, our data support previous substantial evidence that CRT works even in patients with very wide QRSd.

For all these reasons, we believe that questioning the value of CRT in patients with LBBB with QRSd ≥178 ms based on results presented by Sassone et al may not be entirely appropriate. We would like to caution medical public against deferring very wide LBBB patients from CRT. Certainly, rigorous studies on this topic are needed.